Hi,
So we have found how to obtain the correct solution! The main point is to reduce tolerance, after that the problem can be solved even with a coarse mesh. Specifically, with the line:
two_d_model = run_simulation_2d(1e8, 1e-14, 1e1, “2d_monoblock”, export_xdmf=True, mumps_solver=True)
the result for retention is like this:
While increasing tolerance to
two_d_model = run_simulation_2d(1e10, 1e-12, 1e1, “2d_monoblock”, export_xdmf=True, mumps_solver=True)
results in the following:
retention just stops to increase at some point in time, and this time is getting smaller with further increasing of the tolerance.